AI Chatbots and Conversational Commerce · Chatbot MILABOT: Chatbot developed by the Montreal...

Kecerdasan Bisnis Terapan

AI Chatbots and

Conversational Commerce

HusniLab. Riset JTIF UTM

Sumber awal: http://mail.tku.edu.tw/myday/teaching/1071/BI/1071BI12_Business_Intelligence.pptx

Business Intelligence (BI)

2

Introduction to BI and Data Science

Descriptive Analytics

Predictive Analytics

Prescriptive Analytics

1

3

5

4

Big Data Analytics

2

6 Future Trends

AI Chatbots and


3

Outline• AI Chatbots

• Conversational Commerce

• Bot Platform Ecosystem

4

AI and

Cognitive Computing

5Source: http://research.ibm.com/cognitive-computing/

Artificial Intelligence (A.I.) Timeline

6Source: https://digitalintelligencetoday.com/artificial-intelligence-timeline-infographic-from-eliza-to-tay-and-beyond/

Chatbots: Evolution of UI/UX

7Source: https://bbvaopen4u.com/en/actualidad/want-know-how-build-conversational-chatbot-here-are-some-tools

AI Chatbot for Conversational

Commerce8

ChatbotDialogue SystemIntelligent Agent

9

Chatbot

10Source: https://www.mdsdecoded.com/blog/the-rise-of-chatbots/

Dialogue System

11Source: Serban, I. V., Lowe, R., Charlin, L., & Pineau, J. (2015). A survey of available corpora for building data-driven dialogue systems. arXiv

preprint arXiv:1512.05742.

Can machines

think?(Alan Turing ,1950)

12Source: Cahn, Jack. "CHATBOT: Architecture, Design, & Development."

PhD diss., University of Pennsylvania, 2017.

Chatbot“online human-computer

dialog systemwith

natural language.”

13Source: Cahn, Jack. "CHATBOT: Architecture, Design, & Development."

PhD diss., University of Pennsylvania, 2017.

Chatbot Conversation Framework

14Source: https://chatbotslife.com/ultimate-guide-to-leveraging-nlp-machine-learning-for-you-chatbot-531ff2dd870c


15

From E-Commerce

to Conversational Commerce:

Chatbotsand

Virtual Assistants

16Source: http://www.guided-selling.org/from-e-commerce-to-conversational-commerce/

Conversational Commerce: eBay AI Chatbots

17Source: https://www.forbes.com/sites/rachelarthur/2017/07/19/conversational-commerce-ebay-ai-chatbot/

Hotel Chatbot

18Source: https://sdtimes.com/amazon/guest-view-capitalize-amazon-lex-available-general-public/


H&M’s Chatbot on Kik


Uber’s Chatbot on Facebook’s Messenger

Uber’s chatbot on Facebook’s messenger

- one main benefit: it loads much faster than the Uber app

Savings Bot

21Source: https://chatbotsmagazine.com/artificial-intelligence-ai-and-fintech-part-1-7cae1e67dc13

Mastercard Makes Commerce More Conversational

22Source: https://newsroom.mastercard.com/press-releases/mastercard-makes-commerce-more-conversational-with-launch-of-chatbots-for-banks-and-merchants/

https://newsroom.mastercard.com/press-releases/mastercard-makes-commerce-more-conversational-with-launch-of-chatbots-for-banks-and-merchants/

Bot Platform

Ecosystem23

24Source: https://www.oreilly.com/ideas/infographic-the-bot-platform-ecosystem

25Source: https://www.oreilly.com/ideas/infographic-the-bot-platform-ecosystem

26Source: https://venturebeat.com/2016/08/11/introducing-the-bots-landscape-170-companies-4-billion-in-funding-thousands-of-bots/

27Source: https://medium.com/@RecastAI/2017-messenger-bot-landscape-a-public-spreadsheet-gathering-1000-messenger-bots-f017fdb1448a /

The Bot Lifecycle

28Source: https://chatbotsmagazine.com/the-bot-lifecycle-1ff357430db7

ChatbotsBot Maturity Model

29Source: https://www.capgemini.com/2017/04/how-can-chatbots-meet-expectations-introducing-the-bot-maturity/

Customers want to have simpler means to interact with businesses and get faster response to a question or complaint.

Chatbot Architectures

• Information Retrieval based Bot (IR-Bot)

• Task Oriented Bot (Task-Bot)

• Chitchat-Bot (Chatbot)

30

Watson DeepQA Architecture

31Source: Ferrucci, David, Eric Brown, Jennifer Chu-Carroll, James Fan, David Gondek, Aditya A. Kalyanpur, Adam Lally et al.

"Building Watson: An overview of the DeepQA project." AI magazine 31, no. 3 (2010): 59-79.

ALICE and AIML

32Source: http://www.alicebot.org/aiml.html

AIML (Artificial Intelligence Markup Language)

<category>

<pattern>HELLO</pattern>

<template>Hi, I am a robot</template>

</category>

33Source: http://www.alicebot.org/aiml.html


• <aiml>

– the tag that begins and ends an AIML document

• <category>

– the tag that marks a "unit of knowledge" in an Alicebot's knowledge base

• <pattern>

– used to contain a simple pattern that matches what a user may say or type to an Alicebot

• <template>

– contains the response to a user input34Source: http://www.alicebot.org/aiml.html

<category><pattern>WHAT ARE YOU</pattern><template>

<think><set name="topic">Me</set></think>I am the latest result in artificial intelligence,which can reproduce the capabilities of the human brainwith greater speed and accuracy.

</template></category>

35


Source: http://www.alicebot.org/aiml.html

Deep Learning for Dialogues

Intent ClassificationIntent LSTM

LSTM (Long-Short Term Memory)GRU (Gated Recurrent Unit)

36Source: Hakkani-Tür, Dilek, Gokhan Tur, Asli Celikyilmaz, Yun-Nung Chen, Jianfeng Gao, Li Deng, and Ye-Yi Wang. "Multi-domain joint semantic frame parsing using bi-directional

RNN-LSTM." In Proceedings of The 17th Annual Meeting of the International Speech Communication Association. 2016.

Dialogue Utterance

37

An example utterance with annotations of semantic slots in

IOB format (S), domain (D), and intent (I),

B-dir and I-dir denote the director name.

Source: Hakkani-Tür, Dilek, Gokhan Tur, Asli Celikyilmaz, Yun-Nung Chen, Jianfeng Gao, Li Deng, and Ye-Yi Wang. "Multi-domain joint semantic frame parsing using bi-directional


End-to-end Memory Network Model for Multi-turn SLU

38Source: Chen, Yun-Nung, Dilek Hakkani-Tür, Gokhan Tur, Jianfeng Gao, and Li Deng. "End-to-end memory networks with knowledge carryover for

multi-turn spoken language understanding." In Proceedings of Interspeech. 2016.

39Source: Chen, Yun-Nung, Dilek Hakkani-Tür, Gokhan Tur, Jianfeng Gao, and Li Deng. "End-to-end memory networks with knowledge carryover for

multi-turn spoken language understanding." In Proceedings of Interspeech. 2016.

Deep Learning for SLU (Spoken Language Understanding)

40Source: Hakkani-Tür, Dilek, Gokhan Tur, Asli Celikyilmaz, Yun-Nung Chen, Jianfeng Gao, Li Deng, and Ye-Yi Wang. "Multi-domain joint semantic frame parsing using bi-directional


Encoder-decoder model for joint intent detection and slot filling

41Source: Liu, Bing, and Ian Lane. "Attention-Based Recurrent Neural Network Models for Joint Intent Detection and Slot Filling."

arXiv preprint arXiv:1609.01454 (2016).

(a) with no aligned inputs.



(b) with aligned inputs.




(c) with aligned inputs and attention


End-to-End Task-Completion Neural Dialogue Systems

44Source: Li, Xuijun, Yun-Nung Chen, Lihong Li, and Jianfeng Gao. "End-to-end task-completion neural dialogue systems."


Reinforcement learning is used to train all components in an end-to-end fashion

SlotIntent



47

Sample dialogues generated by rule-based and RL agents

Source: Li, Xuijun, Yun-Nung Chen, Lihong Li, and Jianfeng Gao. "End-to-end task-completion neural dialogue systems."


48


Source: Li, Xuijun, Yun-Nung Chen, Lihong Li, and Jianfeng Gao. "End-to-end task-completion neural dialogue systems."


A Deep Reinforcement Learning Chatbot

Iulian V. Serban, Chinnadhurai Sankar, Mathieu Germain, Saizheng Zhang, Zhouhan Lin, Sandeep Subramanian, Taesup Kim, Michael Pieper, Sarath

Chandar, Nan Rosemary Ke, Sai Mudumba, Alexandre de Brebisson Jose M. R. Sotelo, Dendi Suhubdy,

Vincent Michalski, Alexandre Nguyen, Joelle Pineauand Yoshua Bengio

Montreal Institute for Learning Algorithms, Montreal, Quebec, Canada

50Source: Serban, Iulian V., Chinnadhurai Sankar, Mathieu Germain, Saizheng Zhang, Zhouhan Lin, Sandeep Subramanian, Taesup Kim et al.

"A Deep Reinforcement Learning Chatbot." arXiv preprint arXiv:1709.02349 (2017).

A Deep Reinforcement Learning Chatbot

MILABOT: Chatbot developed by the

Montreal Institute for Learning Algorithms (MILA)

for the Amazon Alexa Prize competition



MILABOT Dialogue manager control flow



1 2 3

Q: “What is your name?”R: "I am an Alexa Prize Socialbo”

MILABOT Computational graph

for scoring model



model selection policies based on both action-value function and stochastic policy parametrizations

Facebook AI Research : bAbI Project

• The (20) QA bAbI tasks

• The (6) dialog bAbI tasks

• The Children’s Book Test

• The Movie Dialog dataset

• The WikiMovies dataset

• The Dialog-based Language Learning dataset

• The SimpleQuestions dataset

55Source: https://research.fb.com/projects/babi/

Facebook bAbI QA Datasets1 Mary moved to the bathroom.

2 John went to the hallway.

3 Where is Mary? bathroom 1

4 Daniel went back to the hallway.

5 Sandra moved to the garden.

6 Where is Daniel? hallway 4

7 John moved to the office.

8 Sandra journeyed to the bathroom.

9 Where is Daniel? hallway 4

10 Mary moved to the hallway.

11 Daniel travelled to the office.

12 Where is Daniel? office 11

13 John went back to the garden.

14 John moved to the bedroom.

15 Where is Sandra? bathroom 8

1 Sandra travelled to the office.

2 Sandra went to the bathroom.

3 Where is Sandra? bathroom 2


Facebook bAbI QA Datasets

57Source: Weston, Jason, Antoine Bordes, Sumit Chopra, Alexander M. Rush, Bart van Merriënboer, Armand Joulin, and Tomas Mikolov. "Towards

AI-complete question answering: A set of prerequisite toy tasks." arXiv preprint arXiv:1502.05698 (2015).

Learning End-to-End Goal-Oriented Dialog

60

1 hi hello what can i help you with today

2 can you make a restaurant reservation with italian cuisine for six people in a cheap price range i'm on it

3 <SILENCE> where should it be

4 rome please ok let me look into some options for you

5 <SILENCE> api_call italian rome six cheap


Facebook bAbI Dialogue Datasets

62Source: Bordes, Antoine, and Jason Weston. "Learning End-to-End Goal-Oriented Dialog." arXiv preprint arXiv:1605.07683 (2016).

The Dialog bAbI Tasks


Short Text

Conversation(STC)

66

Short Text Conversation Task (STC-3)

Chinese Emotional Conversation Generation (CECG) Subtask

67Source: http://coai.cs.tsinghua.edu.cn/hml/challenge.html

http://coai.cs.tsinghua.edu.cn/hml/challenge.html

NTCIR Short Text ConversationSTC-1, STC-2, STC-3

68Source: https://waseda.app.box.com/v/STC3atNTCIR-14

https://waseda.app.box.com/v/STC3atNTCIR-14

Short Text Conversation (NTCIR-13 STC2)Retrieval-based

69Source: http://ntcirstc.noahlab.com.hk/STC2/stc-cn.htm

http://ntcirstc.noahlab.com.hk/STC2/stc-cn.htm

Short Text Conversation (NTCIR-13 STC2)

Generation-based

70Source: http://ntcirstc.noahlab.com.hk/STC2/stc-cn.htm

http://ntcirstc.noahlab.com.hk/STC2/stc-cn.htm

Short Text Conversation(STC-3)

• Emotional Conversation Generation

• Dialogue Quality

• Nugget Detection subtasks using Chinese and English dialogue data

71Source: http://research.nii.ac.jp/ntcir/ntcir-14/tasks.html

http://research.nii.ac.jp/ntcir/ntcir-14/tasks.html

NTCIR-14 Short Text Conversation

Task (STC-3)

• Three new subtasks

– Chinese Emotional Conversation Generation (CECG)

– Dialogue Quality (DQ)(for Chinese and English)

– Nugget Detection (ND)(for Chinese and English)

72Source: http://research.nii.ac.jp/ntcir/ntcir-14/tasks.html

http://research.nii.ac.jp/ntcir/ntcir-14/tasks.html

Given post: 爱狗还会做饭的男人，最帅了！

The man who cooks and loves dogs is very handsome!

Response Chinese / Translated EnglishEmotion

ClassCoherence

and FluencyEmotion

Consistency Label

Response 1 会做饭的男人是很帅的啊。The man who cooks is handsome.

喜欢Like

Yes Yes 2

Response 2 哈哈，我也觉得。Haha, I feel the same way.

喜欢Like

Yes No 1

Response 3 这是哪部电影里的？Which movie is this from?

厌恶Disgust

No Yes 0

Response 4 哈哈,你也是。Haha, the same to you.

喜欢Like

No No 0

Response 5 我爱你会。I love you, will you.

厌恶Disgust

No Yes 0

Response 6 这是同主义同的道！This is the same way of the same doctrine!

厌恶Disgust

No No 0

73Source: http://coai.cs.tsinghua.edu.cn/hml/challenge/evaluation_metric/

http://coai.cs.tsinghua.edu.cn/hml/challenge/evaluation_metric/

NLPCC Emotion Classification Challenge

6 classes (64%)

• 0: Other

• 1: Like

• 2: Sadness

• 3: Disgust

• 4: Anger

• 5: Happiness

74Source: http://coai.cs.tsinghua.edu.cn/hml/challenge/dataset_description/

http://coai.cs.tsinghua.edu.cn/hml/challenge/dataset_description/

Chinese Emotional Conversation Generation (CECG)

Post(Given)

Emotion Category (Given)

Response(to be generated)

爱狗还会做饭的男人，最帅了！The man who cooks and loves dogs is very handsome!

喜欢Like

会做饭的男人是很帅的啊。The man who cooks is handsome.

爸爸，明天是六一啦，带我出去玩吧！Daddy, tomorrow is June 1th Children‘s Day. Please take me out to play!

喜欢Like

好，一定带。Yes, of course.

被特别在乎的人忽略，会很难过，而装做不在乎，会更难过。It's sad to be ignored by someone that you really care, and it's harder to pretend that you don't care.

悲伤Sadness

因为在乎，所以在乎。I am concerned with it, because I care about it.

冰天雪地的哈尔滨！你要冻死我咋地？！？！Harbin is in the snow! Are you going to freeze me to death?!?!

喜欢Like

注意保暖，小心别感冒!Keep warm. Watch out for getting cold!

本来想学一把沧桑，结果令我更忧伤。I wanted to learn the vicissitudes of life, but I became sadder.

悲伤Sadness

这是一个悲伤的故事。It a sad story.

今天一整天都在海边训练。虽然很累，但还是很开心的！I have been training at the seaside all day. Though very tired, I still very happy!

快乐Happiness

是的呢，开心！Yeah, happy!

75Source: http://coai.cs.tsinghua.edu.cn/hml/challenge/task_definition/

http://coai.cs.tsinghua.edu.cn/hml/challenge/task_definition/

Chinese Emotional Conversation Generation

(CECG)Dataset

• 1,110,000 Weibo post-response pairs

– [[[post,post_label],[response,response_label]],[[post,post_label],[response,response_label]],...].

76Source: http://coai.cs.tsinghua.edu.cn/hml/challenge/dataset_description/

http://coai.cs.tsinghua.edu.cn/hml/challenge/dataset_description/

Fluency judgement on responses with repetitive words

77

Source: Huang, Minlie, Zuoxian Ye, and Hao Zhou. "Overview of the NLPCC 2017 Shared Task: Emotion Generation Challenge." In National CCF Conference on Natural Language Processing and Chinese Computing (NLPCC), pp. 926-936. Springer, Cham, 2017.

Sample responses generated by Seq2Seq and ECM

(Emotional Chatting Machine)

78Source: Zhou, Hao, Minlie Huang, Tianyang Zhang, Xiaoyan Zhu, and Bing Liu. "Emotional chatting machine:

emotional conversation generation with internal and external memory." arXiv preprint arXiv:1704.01074 (2017).

Emotional Short Text Conversation(ESTC)

Dataset



Conversations with/without considering emotionEmotional Chatting Machine (ECM)

• User: Worst day ever. I arrived late because of the traffic.

1. Basic Seq2Seq: You were late.

2. ECM (Like): I am always here to support you.

3. ECM (Happy): Keep smiling! Things will get better.

4. ECM (Sad): It’s depressing.

5. ECM (Disgust): Sometimes life just sucks.

6. ECM (Angry): The traffic is too bad!



Overview of Emotional Chatting Machine (ECM)



Data flow of the decoder with an internal memory



Data flow of the decoder with an external memory



Chinese Emotional Conversation Generation

(CECG)Evaluation Metric

• Emotion Consistency

– whether the emotion class of a generated response is the same as the pre-specified class.

• Coherence

– whether the response is appropriate in terms of both logically coherent and topic relevant content.

• Fluency

– whether the response is fluent in grammar and acceptable as a natural language response.



Chinese Emotional Conversation Generation (CECG)Evaluation Metric

IF Coherence and Fluency

IF Emotion Consistency

LABEL 2

ELSE

LABEL 1

ELSE

LABEL 0



Sequence-to-sequence Learning with Attention for Generation-based STC

90Source: Yu, Kai, Zijian Zhao, Xueyang Wu, Hongtao Lin, and Xuan Liu. "Rich Short Text Conversation Using Semantic

Key Controlled Sequence Generation." IEEE/ACM Transactions on Audio, Speech, and Language Processing (2018).

General Framework of Controllable Short-Text-Conversation Generation

with External Memory



Controllable Short Text Conversation Examples



Comments Generated Using Different Semantic key Mapping Methods



Generated Responses of Knowledge Introduction by External Memory



How to Build Chatbots

95Source: Igor Bobriakov (2018), https://activewizards.com/blog/a-comparative-analysis-of-chatbots-apis/

Chatbot Frameworks and AI Services

• Bot Frameworks

– Botkit

– Microsoft Bot Framework

– Rasa NLU

• AI Services

– Wit.ai

– api.ai

– LUIS.ai

– IBM Watson96Source: Igor Bobriakov (2018), https://activewizards.com/blog/a-comparative-analysis-of-chatbots-apis/

Chatbot Frameworks


RasaConversational AI

99

Rasa: Conversational AI

100https://rasa.com/

https://rasa.com/

Rasa PlatformRasa Stack

101https://rasa.com/

https://rasa.com/

Rasa Core High-Level Architecture

102Source: https://rasa.com/docs/core/architecture/

Rasa the OSS to build conversational software with ML

103Source: Justina Petraitytė (2018), Deprecating the state machine: building conversational AI with the Rasa stack, PyData Berlin 2018.

https://github.com/RasaHQ/rasa-workshop-pydata-berlin

https://github.com/RasaHQ/rasa-workshop-pydata-berlin

Rasa NLU: Natural Language Understanding


Rasa Core: Dialogue Handling


Rasa Core: Dialogue Training


Dialogflow

108https://dialogflow.com/

https://dialogflow.com/

Learning Semantic Textual Similarity from Conversations

109Source: Yinfei Yang, Steve Yuan, Daniel Cer, Sheng-yi Kong, Noah Constant, Petr Pilar, Heming Ge, Yun-Hsuan Sung, Brian Strope, and

Ray Kurzweil (2018). "Learning Semantic Textual Similarity from Conversations." arXiv preprint arXiv:1804.07754.

TF-Hub ModulesSentence Embedding

Universal Sentence Encoder

110https://tfhub.dev/

https://tfhub.dev/

Semantic Similarity with TF-HubUniversity Sentence Encoder

111https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/semantic_similarity_with_tf_hub_universal_encoder.ipynb

https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/semantic_similarity_with_tf_hub_universal_encoder.ipynb

Semantic Textual Similarity

112https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/semantic_similarity_with_tf_hub_universal_encoder.ipynb

https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/semantic_similarity_with_tf_hub_universal_encoder.ipynb

anaGoSequence Labeling (NER)

113https://github.com/Hironsan/anago

https://github.com/Hironsan/anago

GRAM-CNNBioNER

• GRAM-CNN is a novel end-to-end approach for biomedical NER tasks. To automatically label a word, this method uses the local information around the word. Therefore, the GRAM-CNN method doesn't require any specific knowledge or feature engineering and can be theoretically applied to all existing NER problems.

• The GRAM-CNN approach was evaluated on three well-known biomedical datasets containing different BioNER entities. It obtained an F1-score of 87.38% on the Biocreative II dataset, 86.65% on the NCBI dataset, and 72.57% on the JNLPBA dataset. Those results put GRAM-CNN in the lead of the biological NER methods.

• Pre-trained embedding are from: – https://github.com/cambridgeltl/BioNLP-2016

114Source: Zhu, Qile, Xiaolin Li, Ana Conesa, and Cécile Pereira. "GRAM-CNN: a deep learning approach with local context for named

entity recognition in biomedical text." Bioinformatics 34, no. 9 (2017): 1547-1554. https://github.com/valdersoul/GRAM-CNN

https://github.com/valdersoul/GRAM-CNN

Summary• AI Chatbots

• Conversational Commerce

• Bot Platform Ecosystem

115

References• Yun-Nung Chen, Dilek Hakkani-Tür, Gokhan Tur, Jianfeng Gao, and Li Deng (2016), "End-to-end memory networks with knowledge

carryover for multi-turn spoken language understanding." In Proceedings of Interspeech.

• Hakkani-Tür, Dilek, Gokhan Tur, Asli Celikyilmaz, Yun-Nung Chen, Jianfeng Gao, Li Deng, and Ye-Yi Wang. "Multi-domain jointsemantic frame parsing using bi-directional RNN-LSTM." In Proceedings of The 17th Annual Meeting of the International SpeechCommunication Association. 2016.

• Jeongkyu Shin (2016), Building AI Chat bot with Python 3 and TensorFlow, PyCon APAC 2016.

• Ankit Kumar, Ozan Irsoy, Jonathan Su, James Bradbury, Robert English, Brian Pierce, Peter Ondruska, Ishaan Gulrajani, and RichardSocher (2015). "Ask me anything: Dynamic memory networks for natural language processing." arXiv preprint arXiv:1506.07285.

• Antoine Bordes and Jason Weston (2016), "Learning End-to-End Goal-Oriented Dialog." arXiv preprint arXiv:1605.07683.

• Jason Weston, Antoine Bordes, Sumit Chopra, Alexander M. Rush, Bart van Merriënboer, Armand Joulin, and Tomas Mikolov (2015),"Towards AI-complete question answering: A set of prerequisite toy tasks." arXiv preprint arXiv:1502.05698.

• Iulian V. Serban, Chinnadhurai Sankar, Mathieu Germain, Saizheng Zhang, Zhouhan Lin, Sandeep Subramanian, Taesup Kim et al(2017). "A Deep Reinforcement Learning Chatbot." arXiv preprint arXiv:1709.02349.

• Xuijun Li, Yun-Nung Chen, Lihong Li, and Jianfeng Gao (2017), "End-to-end task-completion neural dialogue systems." arXiv preprintarXiv:1703.01008.

• Jack Cahn (2017), "CHATBOT: Architecture, Design, & Development." PhD diss., University of Pennsylvania.

• Ferrucci, David, Eric Brown, Jennifer Chu-Carroll, James Fan, David Gondek, Aditya A. Kalyanpur, Adam Lally et al.(2010) ". BuildingWatson: An overview of the DeepQA project" AI magazine 31, no. 3 (2010): 59-79.

• Kato, Makoto P., and Yiqun Liu (2017), "Overview of NTCIR-13." In Proceedings of the 13th NTCIR Conference. 2017.

• Huang, Minlie, Zuoxian Ye, and Hao Zhou (2017). "Overview of the NLPCC 2017 Shared Task: Emotion Generation Challenge."In National CCF Conference on Natural Language Processing and Chinese Computing (NLPCC), pp. 926-936. Springer, Cham, 2017.

• Zhou, Hao, Minlie Huang, Tianyang Zhang, Xiaoyan Zhu, and Bing Liu. (2017) "Emotional chatting machine: emotional conversationgeneration with internal and external memory." arXiv preprint arXiv:1704.01074.

• Yu, Kai, Zijian Zhao, Xueyang Wu, Hongtao Lin, and Xuan Liu. (2018)"Rich Short Text Conversation Using Semantic Key ControlledSequence Generation." IEEE/ACM Transactions on Audio, Speech, and Language Processing.

• Justina Petraitytė (2018), Deprecating the state machine: building conversational AI with the Rasa stack, PyData Berlin 2018.

• Yinfei Yang, Steve Yuan, Daniel Cer, Sheng-yi Kong, Noah Constant, Petr Pilar, Heming Ge, Yun-Hsuan Sung, Brian Strope, and RayKurzweil (2018). "Learning Semantic Textual Similarity from Conversations." arXiv preprint arXiv:1804.07754.

116

AI Chatbots and Conversational Commerce · Chatbot MILABOT: Chatbot developed by the Montreal...

Documents

Transcript of AI Chatbots and Conversational Commerce · Chatbot MILABOT: Chatbot developed by the Montreal...